Messages in this thread | | | Date | Tue, 2 Dec 2014 09:59:49 +0100 | From | Michal Hocko <> | Subject | Re: oom && coredump |
| |
On Thu 27-11-14 18:47:47, Oleg Nesterov wrote: > On 11/27, Michal Hocko wrote: [...] > > Why is the ptrace check wrong? > > It was added in reply to exploit I sent. But: > > - It doesn't (and can't) really work, it can only detect this particular > case and the same exploit still blocks oom-killer with the minimal > modifications. > > - Once again, this has nothing to do with ptrace. That exploit used > ptrace only to control (freeze) the coredumping process, the coredumping > can "hang" because of other reasons. > > - It is no longer needed after this patch, the coredumping process will > be killed.
OK, I guess I am seeing it now. Thanks for the clarification!
[...] -- Michal Hocko SUSE Labs
| |